1. Layer 1 Connectivity Issues Part 1
We are now in the troubleshooting section and the first part is troubleshooting layer one connectivity. Our first example consists of a pair of big IP device and big IP. One is connected to the authenticated switch that is also connected to three servers. We have another switch, the switch to is connected into the second big IP device and another three servers. Now this big IP is designed for high availability and it is a high availability pair. One is active and one is standby. Now the status here is you are just testing the failover and again this is the big IP one which is active and this is the standby. So failover is not working properly from the big IP active standby pair. Now you check some status and you thought that it is fine.
You don’t see anything wrong with the configuration and you come up with the tmSh output. You run a show interface under tmSh net module and this is the output. As you can see one one interface is up and maybe this is connected to another switches or another switch because this is for the external network. One is interface also up. You don’t see any problems. You see bits in, bits out packaging and out. But you see this interface one three doesn’t have a status of up, it has a status of miss. What is miss anyway? And what should we check if we see this kind of status? I’m here in Ask F five, page K 14538 and look at the title big IP 2004 Thousand series Platforms fails to detect a specific module.
Now if I scroll down you will see the same status as what we have in our big IP one device. The status is missed. And if you also check the logs under bar log Ltm you will see this message the request media for the interface. Well, in our case, one three is invalid. Now, what is this supposed to mean is if I scroll up here’s the impact if you use this specific module, one gigabit LX Fiber Sfp module with its model number, it will not function on both Big IP 2004 Thousand series platform.Now this is already a known issue and for this article it already provides us a resolution. One of the fixes is just upgrade to this version eleven four one. Or if you’re still experiencing this kind of issue, you can upgrade your big IP device.
Under this link it says most recent version of the F five software. Or you can also upgrade it or add hot fixes. As you can see, the big IP issue is resolved by again upgrading software. Now another work around is you may be able to mitigate the issue by changing the module to one gig SR or ten gig LR Fiber Sfp. We have another article related to the output and this is K one three 1116. It says Interface may be reported as unpopulated or missed. Exactly what we see in our show interface net output and the symptoms by the way is not only named as Miss well, if you use your GUI it doesn’t show you the same name, it will show you unpopulated. Miss will only be seen if you do TMSA show net interface and from the old big IP cli which is also known as the big pipe cli, it doesn’t also view as myth but the status is Ms.
Now, there’s no solution for this and the article is also very old like the previous one. But this issue occurs one of the following conditions the first one is related to VIP prion blade and as we mentioned Vipreon is the modular the chassis base f five big IP appliance and this occurs when the Vipreon blade cluster status is changed to disabled. I don’t think this will be asked in the exam or you will see related options one is network failover feature is not appropriately configured or it’s not working at all. Now for your exam you may see this kind of questions what is missed status? And in our discussion it has two possible related issue one can be module related issue a module is not detected like what we see in the first article and another option is possibly related to network failover feature. But for your F 5101 exam you have to read carefully what are the options? Compare them and choose the best answer. Bye.
2. Layer 1 Connectivity Issues Part 2
In our second example we have a different topology. At this time we only have a single big IP appliance and our big IP appliance is connected to a switch and this switch is also connected to three servers. Now our issue is you are experiencing connection loss and decided to run show interface from tmSh underneath module. You verify the switch is up and running. So this device, this device, this switch it’s up. You see some lights that are blinking. It’s verified that it’s up and running. Now when you run the show interface command under tmSh net module you see that one two interface of the big IP is down down status. This one here is down. If you compare it to the other interface one one is up.
This is the interface that is connected to our external network. So it is fine, it’s not the issue. Now the issue is one two interface and it is down. How about the third interface one three? It says uninit or uninitialized. That means you didn’t configure it to be up. There is no Vlan, there’s no self IP address. It’s also not disabled because if you disabled it manually the status would look like disabled because it will tell you it is disabled. So we don’t need to bother the third interface. What we are trying to understand why the interface one two has the status of that. We already know that this switch is up and running. So the question would be what are the possible issue? Well, obviously this is a layer one issue.
There are many different possible issues that we can figure out. Since the switch is already up it’s not possible, it’s almost not impossible or the chances it’s very slim that the switch itself, the hardware has some hardware issue. Okay, maybe not. Now what you can check is this interface, this specific interface day E one. Why? Well, maybe someone accidentally shut it down manually or maybe it’s you, right? So you have to verify the status of this interface. It can be manually shut down by someone. It can also be a cable problem. Sometimes this happens you thought the cable that you bought is in good quality but all of a sudden it starts not working. So one issue is configuration issue or someone manually set it to down or administratively down. Config down can be a bad cable issue.
This is one, this is two. Another issue is also configuration related issue and still related to E one and can be also related to E one Two, which is the interface of our Fib IP device. Now, what’s supposed to be related for the configuration and what can cause this status to be down? We’ve already talked about a little about speed and duplex mismatch. Sometimes if two devices switch or router or big IP or firewall if their settings doesn’t match, let’s say speed. One is one gig, the other is ten gig. Sometimes the interface will go down because of the speed setting mismatch. Also it can be a duplex setting. So I will specify here speed and it’s also possible duplex. So this is three and four. Okay, so it can be duplex setting as well. So your switch may be set to half duplex.
Someone configured that or maybe you accidentally configured that, e one interface to be half duplex or maybe you’re experimenting and then you forgot to put it back to full duplex and your VG IP device already by default set to full duplex. And sometimes in the configuration you cannot even change it to half duplex. So it can be configuration related issue mismatching, the duplex setting or the speed setting. So that is the third option. So these are the possible configuration related issue and hardware related issue like the by cable or the interface itself is set to down. Or it can also be a faulty interface. This can be faulty interface. I can actually add this is three and four. This is five faulty interface. Okay, so again in your exam you have to compare the questions and choose the best answer.
3. Layer 1 Connectivity Issues Part 3
Our third example is not only related to physical connectivity but also related to applications. Here is our setup. We have three servers and our three servers is listening or are listening to port 80. This servers is configured in port pool, we call it Http pool and this Http pull is associated to our Http Vs with an IP address of 1010 1100 listening to port 80. We have a client here and the client is requesting Http connection. Now, it is reported that the clients are experiencing latency when connecting to these Http applications. And you as the Fibit specialist you are trying to troubleshoot and you’ve seen some outputs. You see the output under Show virtual server under tmSh Ltm module and under Http underscore Vs you see that the availability is available, the state is enabled.
You see traffic is coming in, bits in and out, packets in and out. Another output that you verify is the pool under pool HDT pool you see also the availability and state and you see how many available members under this Http pool. We also have another output, this is under tmSh net module Show interface. As you see both interface one one and one two are up. What seems to be a problem now in the exam you may see many different options and sometimes they are not even related to each other. You may see something like virtual server is not available or is offline or something wrong with a virtual server. We’ve already verified that the Vs is available and enabled and you see packets coming in, bits coming in and out.
So if the option is related to virtual servers might be not the best answer. You may also see pull or node related question or excuse me, pull or node related answer. We’ve already verified that the pool Http pool is available and this is not related to any offline or not offline. It’s not related to pool members or offline as well. Because again, even if you only have one pool member and the rest are offline, your pool will still become available. All you need is one pool member for your application to work. So it’s not related to node or pool member. And the reason why you still don’t have a node output here but it is not related to node also might not be.
This is a high probability because you see bits in and out, packets in and out to your Http pool. Now we have another output and again one two and one interface both are up. One three don’t mind because we’re not using it anyway. But look carefully, there are bits in and out, packets in and out, but there are also drops incrementing. If you compare the options virtual server pools, full numbers, nodes and this interface, it might be a layer one issue and this can be related to drops. And there are many reasons why there are drops incrementing in one or more interface. But again you have to compare the options and from there you will be able to choose the correct answer.
4. Layer 1 Connectivity Issues Part 4
Here’s our Ford example under layer one connectivity issue. Now, as you can see our switch is connected to a pair of servers and a pair of PCs. Now in the exam it’s not always the question related to big IP. This case we are talking about networking as a genetic technology and our device or our network device is a switch. It’s not a big IP but a switch layer to switch and can be made by any vendor. Now, the issue is one PC and a server has or experiencing slowness when it comes to communication. OK? And upon checking the status of the interface e One and E Four of the switch, the switch we have E one interface, an E Four interface connected to the first PC on the first well, this is actually the second server.
You see a number of collisions continuously incrementing what may cause this collision. First off, in our big IP device you will not see any collision statistics or even if there is, it’s not under Show interface. It may be somewhere else, maybe you need to add detail but it is more popular or more well known to see this in a layer to device. And I’m going to explain why in a bit. This switch, if you do in some popular switch vendor you just do Show interface plus the interface name you will see down below you will see the collision number and it may increment depending on the status of your switch or your endpoint device.
But let’s figure it out, what’s happening first. This interface has settings, right? We have layer two settings. We have the Vlans, we have the Tag or 802 one configuration. These are related to layer two issues. And if you talk about layer two issues especially for the Vlan and tagging, most of the time the result will be the interface would go down. Most of the time. It’s not always, but most of the time. Now, if you see latency or slowness or collision this may be related to layer one but I’m pretty sure if you see collisions it’s high probability is layer one. And when you’re talking about layer one this can be related to physical connectivity like cables. It can be also interface malfunctioning, this interface or this interface.
So if you have options related to something interface, physical interface related it’s also possible but collision by concept it is also related to duplex settings. Now this interfaces, the E One, E Four ideally should be set to form duplex but in some switches this is configured auto. Now, the problem with the default auto, sometimes it doesn’t successfully detecting the interface configuration of the neighboring device, in this case the server or the PC. So sometimes it goes down, sometimes one is full, one is half. So let’s say this is half, this is full, this is half and this is full. Or sometimes someone manually set this interface to half duplex.
Now, if one interface in this case is E one and the interface of the PC is full half and the other is full. It is highly possible, or at least by concept, it will do a lot of collisions. Same with this interface, the E 114 connected to the second server here, second server. So collisions again is very much related to duplex. And the reason why we didn’t use big IP because in many versions of big IP it is already by default set to full duplex and you don’t have the ability or privilege to set it to half. And that’s why our example is a layer to switch. Now, in your exam, if it’s related to layer one like half duplex or bad cable or faulty interface, that’s probably the best answer.
You may see other options related to applications or configuration objects such as false virtual servers. Or maybe they will say maybe the Http applications or Ssh application of the server is not running. That is far from the answer because we are troubleshooting the lowest part of our OSI model, which is layer one and the closest will be layer two. Related configuration and issues. But if the options is related to the application, forget about it. Application slowness most related to the lower layers, layer one and layer two. When we talked about switching and switched devices, we’ve already reduced collision and we talk about collision domains very minimal. Now, what we’re going to focus is collisions and before the switches.
Our network device is a layer one or operating in layer one. This is what we call the hubs. And before switches, if you connect your hosts, your servers, your PCs, your mainframes using hubs and even if it’s a small network or a larger network, you have one large collision domains. And before we talk about collisions and collision domains, I would like to introduce the two main components of the switch. And this is how it determines how the switch perform and operates the data traffic to our network. Now let’s add it here. First switch reduces collisions and I’m also going to add two components. Okay, I will also add a switch and on this interface I will specify the first component.
The first component is the speed. What is the speed? This is how the switch determines how fast your port can send and receive data and the configuration options. On a legacy switches we have ten megabytes or 100 megabits, not megabytes excuse me, megabits of traffic. So I’m going to specify M. Now, in the current switches it is very common to have at least one gig of port. Some are ten gig and some if your company has a budget, it can have up to 100 gig of interface. So it depends on your design. It can be one of these options. Now again, speed. This determines how fast your host or your connected device can send or can receive in the specific port. Now, the second component is what we call duplex.
Now, before we explain duplex, I would like to add a host here, and I’m going to add duplex here. And duplex has two options. Unlike speed, it may vary from the slower 110 MB than 100 MB. Some of these are not available anymore, but it increases every time we have new generations of switches. Now, the duplex has only two modes. We have the half duplex and we have the full duplex. Now, this is how it works. The half duplex. Yes, the PC here can send and can receive data, but not simultaneously simultaneous. Now, the reason why it’s not simultaneous, because it only has one wire for the transmission and for the receiving of data. Now, the question is, can the PC or the host experience collision? Of course, as you can see, we’re only using one wire.
And this wire is either you send or you receive data. Of course, there will be collusion. Is there an avoidance for us to lessen the collision? Yes. This is what we call the CSMA Cmsa CD. This is the collision avoidance. It’s not a collision free protocol. It lessens only the collision. But of course, you will still experience collisions. That is how half duplex works. We also have the full duplex. Now, the full duplex is a bit different because the host here can receive and can send data in a dedicated port or a dedicated wire. Not port, because there’s only one port, dedicated wire. And the good thing about this is you will not experience collision because, again, dedicated port for transmission or send and another dedicated port here to receive.
Now, the reason in our previous example why we are receiving collisions if we use half duplex, because we are only using one wire, and the chances the transmission and the reception will collide at the higher chances, okay? Now, in a switching environment, you will never use half duplex. Maybe by accident, maybe by testing, or maybe for an incorrect auto negotiation. But definitely you will not do this because it’s not something that you would like to experience in your network. Who wants to experience collision? Okay? So again, if you’re using switch, by default, it’s already using full duplex in most cases, most configuration.
5. Layer 1 Connectivity Issues Part 5
Here are the common layer one issues. Now the first two here is related to Cabling. The second one is bad cable if the cable is really faulty and the first one is cable specification. If your cable doesn’t have the correct specifications now you may experience a status of interface down or you may get an error or FCS error on your interface statistics. We also have incorrect media. So your device not only fib IP but other network devices. Well, if they don’t detect the module, it can be Sfp plus base or optical module base. If they don’t detect it, you will get a message. We’ve already talked about this in our first example.
But this is also a common layer one issue speed settings, duplex settings this is configuration related or when your device is not properly detecting the duplex or setting of your neighboring device because some of the device this is configured as auto detect maybe your device and the neighbor device is not in the same vendor. The result would be the interface may go down or in our previous example you may experience coalitions as well. And lastly, the connected device is only available now. It can be the connected device is down at all offline there is no power. It is also possible that the connected device is up and running. But that interface where your device is connected directly is just shut down.
